We were talking about US forces already being alerted. Meaning additional USAF forces in guam and kadena, a few (if not several) carrier groups in vicinity, etc. Invasion is imminent, and everyone knows it kind of situation. Still, it may not pay off to keep a CAP for USAF planes. We're not talking about strategical surprise, but a tactical one. When will strike group A strike, from which direction. Then the next day, or a week after that. There'd be like hundreds of strikes in the first week or two, which can not be intercepted in time by USAF, unless they're already in the air and close by. USN is another story.
It is very likely US wouldn't do anything rash or risky until they've accumulated enough forces in the vicinity of taiwan. So, unless US gets to position all the forces it wants where it wants, before China can organize air strikes (unlikely), there's bound to be little US involvement in the initial stages of the conflict. As time passes, US may very well assemble large enough force so even if we disregard the quality gap, the quantity ratio between available US and chinese planes may not be very large.
We don't know if there's enough present infrastructure to hold majority of PLAAF and PLANAF forces within proper reach of Taiwan. Or, if there is, at what cost? It's one thing using a prepared air force base and another thing using a civilian airport, where sortie generation is likely to be a bit less. Even more so for uprepared runways, etc. (same goes for the Taiwanese operating their planes from the highways and so on)
How many chinese planes would be lost even before main US force is assembled? To gain air superiority one has to fight through a large taiwanese force, a potent air defense network and a navy with very decent air defense capabilities. Granted, a big part of those may be negated with chinese navy and various missiles but still, even a half of the mentioned taiwanese forces would do some serious attrition to any attacker. I don't think it's out of the question that several hundred chinese planes are brought down before air superiority is achieved. (and thats without any US help)
In such a high stakes war there'd be tensions and potential threats on other places of the globe as well. PLAAF will most probably have to spare some of its planes near Korea, Afghanistan, India, etc, etc. There may not be more than a thousand combat planes left for operations against Taiwan, and not all would be fighter planes.
On the other hand, when US forces assemble, we may be looking at several hundred of combat planes, every one being more capable than what chinese can throw in the air. With just Kadena air wings and four carriers (lets say US won't wait to assemble max it could, some 7-8 carries) we're looking at over 300 quality fighter planes. And if USN can provide good enough air cover around Kadena, USAF may decide it's safe enough to house F-22s there, which would give additional quality edge.
In a war of attrition, where Chinese are flying CAS and CAP missions around the clock over Taiwan, and US forces choose the time of their strikes - it's actually more likely US would be the one to have numerical advantage in the air, as well as qualitative. It is actually quite possible Chinese would very quickly learn to just run away and pause their CAS missions. Naturally, that may come very handy to the Taiwanese land army, which would then have the upper hand, since the numbers are on their side versus whatever chinese can land/paradrop/whatever in a given period of time. Not to mention that Chinese forces on Taiwan would not be equipped with heaviest of weaponry, while Taiwanese would enjoy using whatever their have in their arsenal.